Wallstreetcn
2024.04.18 00:52
portai
I'm PortAI, I can summarize articles.

"AIGC First Stock" failed to shine on debut

"AIGC First Stock" Chumenwenwen is listed on the Hong Kong Stock Exchange, but investors are not very enthusiastic about it. Despite having a luxurious lineup of investors, the products and operational aspects are not outstanding enough, leading investors to be cautious. Chumenwenwen plans to globally issue 84.568 million shares, accounting for only 5.67% of the total issued share capital, with an IPO price corresponding to a PS ratio of about 11 times in 2022. Huang Lichong believes that the overall downturn in the capital market and overvaluation are the reasons for the less than ideal subscription for Chumenwenwen's listing. The net proceeds raised are also relatively small, with limited impact on the company's development

Author: Huang Yu

Editor: Zhou Zhiyu

After failing to submit to the Hong Kong Stock Exchange for the first time last year, Mobvoi Inc. (referred to in Chinese as "Chumen Wenwen") is once again knocking on the door of the Hong Kong Stock Exchange in an attempt to claim the title of "AIGC's first stock".

On April 16, Mobvoi Inc. officially started its IPO, which will run until the 19th, and is scheduled to be listed on the main board of the Hong Kong Stock Exchange with the stock code "2438" on April 24.

Under the banner of "AIGC's first stock", the future market performance of Mobvoi will to some extent reflect the market's attitude towards most AIGC companies.

According to sources close to Mobvoi, the oversubscription on the first day of its IPO in Hong Kong exceeded 8 times coverage, and in less than two days, it exceeded 13 times coverage. The international allocation portion was oversubscribed on the first day, with both international and Chinese opinion leader institutions actively participating and supporting this issuance.

However, Huang Lichong, Co-founder of Cozong Strategy Management Group, told Wall Street News that based on Mobvoi's first day IPO subscription situation, the enthusiasm of the capital market subscription is not high. Generally, enthusiastic subscriptions can have 30 to 40 times coverage.

This to some extent also reflects that after experiencing the capital frenzy, investors are no longer as enthusiastic about AIGC companies as before.

Mobvoi has a quite luxurious lineup of investors, and its founder is also a tech heavyweight. However, it is slightly lacking in terms of products and operations, which makes investors cautious.

It is reported that Mobvoi plans to globally issue 84.568 million shares (excluding any shares that may be issued based on the over-allotment option), accounting for only 5.67% of the total issued share capital, with 90% for international offering and 10% for public offering in Hong Kong.

Based on the highest offering price of HKD 4.1 per share, Mobvoi's IPO is expected to raise approximately HKD 347 million, with a maximum market value of approximately HKD 6.115 billion, corresponding to a PS ratio of about 11 times the 2022 market value-to-sales ratio (total market value ÷ main business income).

In Huang Lichong's view, the less than ideal subscription for Mobvoi's IPO is partly due to the overall downturn in the capital market, and also due to its own issue of overvaluation.

Furthermore, Huang Lichong stated that the net proceeds of just over 200 million are insignificant for Mobvoi's development of large models, possibly because there is a wager agreement with investors, so they are eager to go public.

Prior to the IPO, Mobvoi's investor lineup already included financial investors and strategic investors such as SIG Haina Asia, Google, Sequoia China, GoerTek, ZhenFund, and Round Optics.

Although it is a concept stock of AIGC, Mobvoi has been established for ten years. Like most AI companies, Mobvoi also has a founder with a strong technical background Founder of AskMakers, Li Zhifei, is a Ph.D. in Computer Science from Johns Hopkins University, former scientist at Google headquarters, expert in natural language processing and artificial intelligence, and one of the main developers of the world's mainstream machine translation open-source software Joshua.

At the beginning, AskMakers mainly focused on selling smart devices. Later in 2017 and 2020, the business expanded to AI enterprise solutions and AIGC solutions.

As early as 2020, before the wave of large AI models arrived, AskMakers developed the general large model "UCLAI". However, due to the difficulty of commercialization in the past, AskMakers stopped investing in large models and instead started to focus on AIGC products, such as the AI voice assistant MoYin Workshop and the AI digital person Qimiao Yuan.

Later, with the explosive development of AIGC, AskMakers began to restart the development of large models and upgraded it to a multimodal large model "Sequence Monkey" in April 2023.

As a startup company, AskMakers now emphasizes not only pursuing computing power and parameter competition, but also following the principle of "production nourishes models, and models adapt to production", placing AI applications at the core.

Currently, AskMakers focuses on generative AI and speech interaction technology, with three main businesses: AIGC solutions, AI enterprise solutions, and smart devices and accessories.

AI software solutions, including AIGC solutions and AI enterprise solutions, have become the main source of revenue for AskMakers. According to the prospectus, revenue from AI software increased from 0.6 billion yuan in 2021 to 3.43 billion yuan in 2023, with the revenue share rising from 15% in 2021 to 67.7% in 2023.

However, overall revenue growth for AskMakers has slowed down in 2023, almost reaching a standstill domestically. The prospectus shows that the revenue for 2021, 2022, and 2023 was 398 million yuan, 500 million yuan, and 507 million yuan, respectively.

In terms of profit, AskMakers recorded losses of approximately 670 million yuan and 803 million yuan in 2022 and 2023, respectively.

Regarding the use of funds raised from this IPO, AskMakers plans to use about 44.3% to expand the scale of "Sequence Monkey" in the next three years.

However, for a startup company, developing large models is not an easy task and requires massive investment.

In March, Li Zhifei mentioned that currently, models are products, and the difference between large models lies in their performance, which is the result of violence. Behind this violence is resources (money, people, data). Therefore, in a situation where technology, competition, and products have not fully converged, the barrier that can sustain continuous violence is resources.

He believes that there may only be two types of winners when competition converges. One is the wealthy, who can sustain violence due to their financial strength. The other is those with longevity, as the cost of the same violence will decrease by ten times each year. As long as they maintain ultra-low power consumption, long battery life, and find the right entry point to participate in the competition in the later stage of convergence, they have a chance to win Obviously, Ask the Way is not among the former, and whether it can become the latter still needs time to observe.

Huang Lichong is not optimistic about small companies developing large models on their own or using open-source large models. He told Wall Street News that a company like Ask the Way, with only 146 million RMB in cash on its books, cannot convince people that they can do well with large language models. If they use their own models for their applications, it is equivalent to castrating themselves.

Previously, Baidu's founder, Robin Li, has also guided AI entrepreneurs multiple times, believing that "developing large models is meaningless, while developing applications has greater opportunities."

Of course, while emphasizing the advantages of developing AI large models on their own, Ask the Way also strives to highlight its own AI application capabilities. In the context of the increasing speed of AI large model development and application iteration, Ask the Way, standing in the spotlight after going public, is destined for a journey full of unknowns, whether they can proceed more steadily and further